Figure 2. The effect of carbon farming practices on soil redox potential. (Light shade= control plot; dark shade = carbon farming plot. Samples taken in July 2021, on the third year of the experiment.)

Redox complements, but does not replace other measures of soil health

The Eh correlated significantly with CO2burst (p<0.001) and explained 36% of variability (R2=0.36). In contrast, the correlations of pH (p=0.06) and clay (p=0.80) with CO2burst were nonsignificant. Organic matter (OM) by itself was not-significant (p > 0.10; R2 = 0.08), but when included in a linear regression with Eh, it was significant (p < 0.001) and together the two variables explained 50% of the variability in CO2burst. This supports earlier findings, that the CO2burst is more related to the availability and quality of organic matter than the quantity (Haney et al., 2012). Redox potential explains the variation in biological activity better than organic matter or texture. It may be, that this is due to the rewetting experiment conditions, where labile carbon compounds are oxidized (eq. 1), and that the redox is not directly related to the biological activity, but to the physicochemical factors controlling CO2 flush from rewetting (Barnard et al., 2020). In any case, when combined with soil OM, the redox potential can explain the majority of the biological activity related to the carbon cycle, making it a promising soil health indicator or a rapid proxy for respiration tests.
However, the redox should not be used to predict CO2burst. Even when including the pH correction to Eh (i.e. hydrogen potential rH2), redox explained only 35% of the variability (R2=0.35) (Figure 3). Reduced conditions co-existed with high biological activity: of the five soils with high biological activity, four had reduced redox status (rH2< 28). The one exception was a very high OM level (14%) pasture site. Thus, reduced conditions were a requisite for high biological activity. However, as also three reduced soils had only moderate biological activity, reduced conditions do not guarantee a high biological activity. Likewise, the CO2 burst was highly variable under oxidized conditions, suggesting that redox cannot replace CO2 burst as a biological soil health test.
The trend of oxidized redox and low biological activity in the lab tests is in direct contrast with field observations, where high oxidation correlates with high respiration (Bartolucci et al., 2021). However, in field conditions, the redox was changing over time (with soil drying) and in our observations the redox changed between samples in an artificial drying-rewetting experiment. In the field experiments, oxygen limited respiration on a rewetted wetland. In the laboratory test, respiration was more likely limited by carbon compound availability or the amount of microbial biomass, similar to the CO2burst (Barnard et al., 2020). In any case, field experiments on drained, low organic matter soils would be needed to estimate how well the lab test redox results correlate with in-field CO2 burst events.